Model selection and estimation in the Gaussian graphical model
The DOI matches its Crossref record; 16 of the 16 references it cites were found in a registry or an index.
Abstract
We propose penalized likelihood methods for estimating the concentration matrix in the Gaussian graphical model. The methods lead to a sparse and shrinkage estimator of the concentration matrix that is positive definite, and thus conduct model selection and estimation simultaneously. The implementation of the methods is nontrivial because of the positive definite constraint on the concentration matrix, but we show that the computation can be done effectively by taking advantage of the efficient maxdet algorithm developed in convex optimization. We propose a BIC-type criterion for the selection of the tuning parameter in the penalized likelihood methods. The connection between our methods and existing methods is illustrated. Simulations and real examples demonstrate the competitive performance of the new methods.
Overview
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
In the citing paper
“These methods have the advantage to be able to recover the off diagonal structure of the covariance matrix.”
From Signal Estimation and Uncertainties Extraction in Terahertz Time-Domain Spectroscopy (2025) · cites this paper as [58]
Citations
Cites
Boyd, S., & Vandenberghe, L. (2004) · doi:10.1017/cbo9780511804441 · Cited by 56,601
Regression shrinkage and selection via the lasso
Tibshirani, R. (1996) · Journal of the Royal Statistical Society Series B: Statistical Methodology · doi:10.1111/j.2517-6161.1996.tb02080.x · Cited by 51,482 · cited 1×
“This is similar to the idea of the lasso in linear regression (Tibshirani, 1996).”
MethodGraphical Models in Applied Multivariate Statistics
Goodall, C. R. (1991) · Technometrics · doi:10.1080/00401706.1991.10484880 · Cited by 1,849
Asymptotics for lasso-type estimators
Fu, W., & Knight, K. (2000) · The Annals of Statistics · doi:10.1214/aos/1015957397 · Cited by 1,276 · cited 2×
“, X(p)) be a p-dimensional random vector following a multivariate normal distribution Np(µ, ) with unknown mean µ and nonsingular covariance matrix .”
BackgroundMethodInfluentialHeuristics of instability and stabilization in model selection
Breiman, L. (1996) · The Annals of Statistics · doi:10.1214/aos/1032181158 · Cited by 1,146 · cited 1×
“As demonstrated by Breiman (1996), the discrete nature of such procedures often leads to instability of the estimator: small changes in the data may result in very different estimates.”
BackgroundBetter Subset Regression Using the Nonnegative Garrote
Breiman, L. (1995) · Technometrics · doi:10.1080/00401706.1995.10484371 · Cited by 1,043 · cited 2×
“As demonstrated by Breiman (1996), the discrete nature of such procedures often leads to instability of the estimator: small changes in the data may result in very different estimates.”
BackgroundMethodIntroduction to Graphical Modelling
Edwards, D. (2000) · Springer Texts in Statistics · doi:10.1007/978-1-4612-0493-0 · Cited by 825
Determinant Maximization with Linear Matrix Inequality Constraints
Vandenberghe, L., Boyd, S., & Wu, S.-P. (1998) · SIAM Journal on Matrix Analysis and Applications · doi:10.1137/s0895479896303430 · Cited by 652 · cited 2×
“To use the algorithm of Vandenberghe et al. (1998), it is also required that Fi, i = 1,. .”
MethodMultivariate Dependencies: Models, Analysis and Interpretation.
Murphy, S. A., Cox, D. R., & Wermouth, N. (1998) · Biometrics · doi:10.2307/2534028 · Cited by 487
Sparsistency and rates of convergence in large covariance matrix estimation
Lam, C., & Fan, J. (2009) · The Annals of Statistics · doi:10.1214/09-aos720 · Cited by 418
Cited by
Items 401–500 of 1,000
Learning linear structural equation models in polynomial time and sample complexityPreprint
Ghoshal, A., & Honorio, J. (2017) · arXiv (Cornell University) · doi:10.48550/arxiv.1707.04673 · Cited by 60
“A popular approach for inverse covariance estimation, under high-dimensional settings, is the 1 -penalized Gaussian maximum likelihood estimate (MLE) studied by [27], [1], and [7], among others.”
BackgroundBilgrau, A. E., Peeters, C. F. W., Eriksen, P. S., Bøgsted, M., & van Wieringen, W. N. (2015) · VBN Forskningsportal (Aalborg Universitet) · doi:10.48550/arxiv.1509.07982 · Cited by 25
“A popular approach is to amend (1) with an (cid:96) 1 -penalty [48, 2, 19, 49].”
BackgroundNetwork-Scale Traffic Modeling and Forecasting with Graphical Lasso
Gao, Y., Sun, S., & Shi, D. (2011) · Lecture notes in computer science · doi:10.1007/978-3-642-21090-7_18 · Cited by 23
“Basing the selected penalty parameter, for finite samples, the probability of error in estimating the graphical model is controlled. More details can be seen in [ 7 ].”
BackgroundMethodInfluentialBayesian Graphical Models for Differential Pathways
Mitra, R., Müller, P., & Ji, Y. (2015) · Bayesian Analysis · doi:10.1214/14-ba931 · Cited by 23
“Related proposals are mostly based on lasso and penalized likelihood techniques (Yuan and Lin, 2007).”
MethodPartial correlation matrix estimation using ridge penalty followed by thresholding and re‐estimation
Ha, M. J., & Sun, W. (2014) · Biometrics · doi:10.1111/biom.12186 · Cited by 28
“If γ = 0, the classical BIC used in Yuan and Lin (2007) was recovered.”
BackgroundMethodVariable selection for high dimensional multivariate outcomes
Sofer, T., Dicker, L. H., & Lin, X. (2013) · Statistica Sinica · doi:10.5705/ss.2013.019 · Cited by 23
“For instance, Schäfer and Strimmer (2005) proposed shrinkage estimators, while Meinshausen and Buhlmann (2006) and Yuan and Lin (2007) proposed Lasso penalized esti- mators for the covariance or precision matrix.”
BackgroundAberrant Brain Connectivity in Schizophrenia Detected via a Fast Gaussian Graphical Model
Zhang, A., Fang, J., Liang, F., Calhoun, V. D., & Wang, Y. (2018) · IEEE Journal of Biomedical and Health Informatics · doi:10.1109/jbhi.2018.2854659 · Cited by 27
“Various solutions have been proposed in the literature, such as nodewise regression [2] and graphical Lasso (gLasso) [3], [4].”
AIC for the non-concave penalized likelihood method
Umezu, Y., Shimizu, Y., Masuda, H., & Ninomiya, Y. (2018) · Annals of the Institute of Statistical Mathematics · doi:10.1007/s10463-018-0649-x · Cited by 23
“…caused by selecting a model based on only one value of λ would be nice, it carries with it a considerable computational cost as in CV. Recently, information criteria without such a problem have been developed (Yuan and Lin 2007; Wang et al. 2007, 2009; Zhang et al. 2010; Fan and Tang 2013).”
BackgroundBenfenati, A., Chouzenoux, É., & Pesquet, J.-C. (2019) · Signal Processing · doi:10.1016/j.sigpro.2019.107417 · Cited by 29
“We can also mention the dual block coordinate ascent method from [3], the SPICE algorithm [30], the gradient projection method in [1], the Refitted CLIME algorithm [31], various algorithms [8] based on Nesterov’s smooth gradient approach [32], ADMM approaches [33], an inexact Newton method [34], and interior point met…”
BackgroundMethodTwo-Sample Inference for High-Dimensional Markov Networks
Kim, B., Liu, S., & Kolar, M. (2021) · Journal of the Royal Statistical Society Series B (Statistical Methodology) · doi:10.1111/rssb.12446 · Cited by 23
“In high- dimensional regimes, consistent estimation requires an assumption of inherent low- dimensionality such as sparsity (Cai et al., 2011; Friedman et al., 2008; Ravikumar et al., 2011; Yuan, 2010; Yuan & Lin, 2007).”
BackgroundAccelerated Graph Learning From Smooth Signals
Saboksayr, S. S., & Mateos, G. (2021) · IEEE Signal Processing Letters · doi:10.1109/lsp.2021.3123459 · Cited by 23
Tests for Gaussian graphical models
Verzélen, N., & Villers, F. (2008) · Computational Statistics & Data Analysis · doi:10.1016/j.csda.2008.09.022 · Cited by 23
“Huang et al. (2006) and Yuan and Lin (2007) extend this method to directly infer the graph by minimising the log-likelihood penalised by the l 1 norm.”
MethodLocal and Global Inference for High Dimensional Nonparanormal Graphical ModelsPreprint
Gu, Q., Cao, Y., Ning, Y., & Liu, H. (2015) · arXiv (Cornell University) · doi:10.48550/arxiv.1502.02347 · Cited by 23
“…an estimate of Σ is obtained, it can be plugged into any precision matrix estimation procedure for the Gaussian graphical model such as Graphical lasso (Yuan and Lin, 2007; Banerjee et al., 2008; Friedman et al., 2008), CLIME (Cai et al., 2011), graphical Dantzig selector (Yuan, 2010),…”
BackgroundMethodInfluentialHigh-dimensional Sparse Precision Matrix Estimation via Sparse Column Inverse Operator ∗Preprint
Liu, W., & Luo, X. Q. (2012) · arXiv (Cornell University) · Cited by 25
miR-155 regulative network in FLT3 mutated acute myeloid leukemia
Salemi, D., Cammarata, G., Agueli, C., Augugliaro, L., Corrado, C., Bica, M. G., Raimondo, S., Marfia, A., Randazzo, V., Dragotto, P., et al. (2015) · Leukemia Research · doi:10.1016/j.leukres.2015.04.017 · Cited by 26
Gene network reconstruction using global-local shrinkage priors
Leday, G. G. R., de Gunst, M. C. M., Kpogbezan, G. B., van der Vaart, A., van Wieringen, W. N., & van de Wiel, M. A. (2017) · The Annals of Applied Statistics · doi:10.1214/16-aoas990 · Cited by 22
“Various solutions, usually based on resampling or cross-validation, have been proposed [Gao et al. (2012), Lian (2011), Meinshausen and Bühlmann (2010), Foygel and Drton (2010), Giraud (2008), Yuan and Lin (2007)].”
MethodDependence in elliptical partial correlation graphs
Rossell, D., & Zwiernik, P. (2021) · Electronic Journal of Statistics · doi:10.1214/21-ejs1891 · Cited by 22
“Solve the graphical LASSO problem (Friedman, Hastie and Tibshirani, 2008; Yuan and Lin, 2007) by minimizing − log detK + tr(SK) + ρ ∑ i =j |Kij | (11) for a given penalty parameter ρ > 0.”
MethodBayesian Uncertainty Estimation for Gaussian Graphical Models and Centrality Indices
Jongerling, J., Epskamp, S., & Williams, D. R. (2022) · Multivariate Behavioral Research · doi:10.1080/00273171.2021.1978054 · Cited by 25
“The sum P j, k /kðjjjkjÞ in Equation (2) can be taken with or without a penalty on the diagonal terms (Friedman et al., 2008; Meinshausen & B€uhlmann, 2006; Rothman et al., 2010; Yuan & Lin, 2007).”
BackgroundHigh-dimensional functional graphical model structure learning via neighborhood selection approach
Zhao, B., Zhai, P. S., Wang, Y. S., & Kolar, M. (2024) · Electronic Journal of Statistics · doi:10.1214/24-ejs2219 · Cited by 34
Sparse reduced-rank regression with covariance estimation
Chen, L., & Huang, J. Z. (2014) · Statistics and Computing · doi:10.1007/s11222-014-9517-6 · Cited by 22
“Following Yuan and Lin (2007), we estimate the precision matrix using the Gaussian likelihood together with a sparsity-inducing L 1 -penalty on all of its entries min where S e is the sample covariance matrix of the errors, w j j is the (cid:7) j (cid:6) , j (cid:8) entry in Ω , and λ 1 is the…”
MethodRegularized multivariate regression models with skew-t error distributions
Chen, L., Pourahmadi, M., & Maadooliat, M. (2014) · Journal of Statistical Planning and Inference · doi:10.1016/j.jspi.2014.02.001 · Cited by 22
“An n p predictor matrix X is also generated with rows drawn independently from Nð0;ΣXÞ, where ðΣXÞij ¼ 0:7ji jj, as in Yuan and Lin (2007) and Peng et al. (2010).”
MethodEfficient Dimensionality Reduction for High-Dimensional Network Estimation
Çelik, S., Logsdon, B. A., & Lee, S. (2014) · International Conference on Machine Learning · Cited by 28
Functional Graphical Models for Manufacturing Process Modeling
Sun, H., Huang, S., & Jin, R. (2017) · IEEE Transactions on Automation Science and Engineering · doi:10.1109/tase.2017.2693398 · Cited by 22
Node-based learning of differential networks from multi-platform gene expression data
Ou-Yang, L., Zhang, X.-F., Wu, M., & Li, X. (2017) · Methods · doi:10.1016/j.ymeth.2017.05.014 · Cited by 23
“To deal with this problem, a number of studies [20, 21, 22] have instead taken a penalized log-likelihood : max Θ n 2 (log det (Θ)− tr (SΘ))− λ||Θ||1, (1) where S is the sample covariance matrix, det (·) is the determinant of a matrix, tr (·) is the trace of a matrix, ||Θ||1 denotes the sum of the absolute values of t…”
BackgroundLatent Variable Time-varying Network Inference
Tomasi, F., Tozzo, V., Salzo, S., & Verri, A. (2018) · ACM SIGKDD Conference on Knowledge Discovery and Data Mining (KDD) · doi:10.1145/3219819.3220121 · Cited by 24
“The graphical modelling problem has been extensively tackled in literature by estimating the precision matrix Θ instead of the co-variance matrix Σ (as, e.g. , [5]) [15, 30, 34].”
MethodAn Expectation Conditional Maximization Approach for Gaussian Graphical Models
Li, Z. R., & McCormick, T. H. (2019) · Journal of Computational and Graphical Statistics · doi:10.1080/10618600.2019.1609976 · Cited by 22
“…and do not exist when n < p , leading to extensive work on algorithms (and their properties) for estimating (e.g., Meinshausen and Bühlmann 2006; Yuan and Lin 2007; Friedman, Hastie, and Tibshirani 2008; Rothman et al. 2008; Friedman, Hastie, and Tibshirani 2010; Cai, Zhang, and Zhou 2010;…”
BackgroundBasic tail and concentration bounds
Wainwright, M. J. (2019) · Cambridge University Press eBooks · doi:10.1017/9781108627771.002 · Cited by 45
Bayesian Joint Modeling of Multiple Brain Functional Networks
Lukemire, J., Kundu, S., Pagnoni, G., & Guo, Y. (2020) · Journal of the American Statistical Association · doi:10.1080/01621459.2020.1796357 · Cited by 29
“The graphical lasso was run independently for each network over a grid of regularization parameter values, and the optimal graph was selected for each network using a BIC criteria as described in Yuan and Lin (2007).”
MethodA transfer learning approach for network modeling
Huang, S., Li, J., Chen, K., Wu, T., Ye, J., Wu, X., & Yao, L. (2012) · IIE Transactions · doi:10.1080/0740817x.2011.649390 · Cited by 22
“(2008) to implement the penalized maximum likelihood approach proposed by Yuan and Lin (2007), which are applicable to high-dimensional problems.”
MethodSemi-varying coefficient models with a diverging number of components
Li, G., Xue, L., & Lian, H. (2011) · Journal of Multivariate Analysis · doi:10.1016/j.jmva.2011.03.010 · Cited by 22
“More recently, investigation of models with a large number of parameters is revived by the incorporation of variable selection using penalization [20,3,28,26,29].”
A boosting approach to structure learning of graphs with and without prior knowledge
Anjum, S., Arnaud Doucet, B., & Holmes, C. C. (2009) · Bioinformatics · doi:10.1093/bioinformatics/btp485 · Cited by 22
“others (Friedman et al ., 2008, Yuan, 2006, Yuan and Lin, 2007), we generated data from a graphical model with sparse Markov structure ) ij = α for | i − j |= 1; ( (cid:2) ) ii = 1 ∀ i ; ( (cid:2) ) ij = 0 otherwise. noise σ = 0 .”
On a Class of Shrinkage Priors for Covariance Matrix Estimation
Wang, H., & Pillai, N. S. (2013) · Journal of Computational and Graphical Statistics · doi:10.1080/10618600.2013.785732 · Cited by 22
“…formulation, there is a rich literature on methods for developing shrinkage estimators via different penalty functions including the graphical lasso models (Yuan & Lin, 2007; Friedman et al., 2008; Roth-man et al., 2008) and the graphical adaptive lasso models (Fan et al., 2009) among many others.”
MethodCovariance Matrix Estimation in Time Series
Wu, W. B., & Han, X. (2012) · Handbook of statistics · doi:10.1016/b978-0-444-53858-1.00008-9 · Cited by 29
“…setting in which one has multiple iid p-variates random variables, the problem has been extensively studied; see Meinshausen and Bühlman (2006), Yuan and Lin (2007), Rothman et al (2009), Bickel and Levina (2008a, 2008b), Cai et al (2010), Lam and Fan (2009) and Ledoit and Wolf (2004) among…”
BackgroundUnderstanding large text corpora via sparse machine learning
Ghaoui, L. E., Pham, V., Li, G., Duong, V.-A., Srivastava, A. N., & Bhaduri, K. (2013) · Statistical Analysis and Data Mining The ASA Data Science Journal · doi:10.1002/sam.11187 · Cited by 22
“One of the main examples of this approach is the so-called sparse covariance selection problem, with a Gaussian assumption on the data (see [38], and related works such as [22, 34, 50, 44, 31, 29]).”
BackgroundIdentifying network ties from panel data: theory and an application to tax competition
de Paula, Á., Rasul, I., & Souza, P. (2026) · doi:10.1920/wp.cem.2019.5519 · Cited by 29
Simultaneous Inference for Pairwise Graphical Models with Generalized Score MatchingPreprint
Yu, M. Y., Gupta, V., & Kolar, M. (2019) · arXiv (Cornell University) · doi:10.48550/arxiv.1905.06261 · Cited by 22
“Global methods typically estimate the precision matrix by maximizing regularized Gaussian log-likelihood (Yuan and Lin, 2007; Rothman et al., 2008; Friedman et al., 2008; d’Aspremont et al., 2008; Ravikumar et al., 2011; Fan et al., 2009; Lam and Fan, 2009), while local methods estimate the graph…”
MethodInfluentialKovács, S., Ruckstuhl, T., Obrist, H., & Bühlmann, P. (2021) · arXiv (Cornell University) · doi:10.48550/arxiv.2101.02148 · Cited by 22
“Theoretical properties of the Graphical Lasso estimator were investigated by Yuan and Lin (2007); Rothman et al. (2008); Lam and Fan (2009) and Ravikumar et al. (2011) (up to a small difference of whether to penalize the diagonal elements of the precision matrix).”
MethodGraphical Models and Dynamic Latent Factors for Modeling Functional Brain Connectivity
Chang, A., Yao, T., & Allen, G. I. (2019) · IEEE Data Science Workshop (DSW) · doi:10.1109/dsw.2019.8755783 · Cited by 25
“Subsequently, we can obtain an estimate of the sparse, static functional connectivity graph Θ̂ by fitting a graphical model, such as the graphical Lasso [14] or neighborhood selection approach [18], to the residuals Ê = X − B̂ after adjusting for estimated dynamic latent effects.”
MethodEstimating spatial covariance using penalised likelihood with weighted L 1 penalty
Zhu, Z., & Liu, Y. (2009) · Journal of nonparametric statistics · doi:10.1080/10485250903023632 · Cited by 21
“Estimation of the precision matrix using penalized likelihood with different penalties has been studied in the context of longitudinal data analysis (Huang et al., 2006; Levina and Zhu, 2006) and graphical models (Meinshusen and Buhlmann, 2006; Yuan and Lin, 2007).”
BackgroundMethodModelling covariance structure in bivariate marginal models for longitudinal data
Xu, J., & MacKenzie, G. (2012) · Biometrika · doi:10.1093/biomet/ass031 · Cited by 21
A systematic review of structured sparse learning
Qiao, L., Zhang, B.-F., Su, J., & Lu, X. (2017) · Frontiers of Information Technology & Electronic Engineering · doi:10.1631/fitee.1601489 · Cited by 24
Variable Selection in Seemingly Unrelated Regressions with Random Predictors
Puelz, D., Hahn, P. R., & Carvalho, C. M. (2016) · Bayesian Analysis · doi:10.1214/17-ba1053 · Cited by 21
Integrating additional knowledge into the estimation of graphical models
Bu, Y., & Lederer, J. (2021) · The International Journal of Biostatistics · doi:10.1515/ijb-2020-0133 · Cited by 21
Learning Oncogenic Pathways from Binary Genomic Instability Data
Wang, P., Chao, D. L., & Hsu, L.-I. (2010) · Biometrics · doi:10.1111/j.1541-0420.2010.01417.x · Cited by 23
“Other approaches include penalized maximum likelihood (e.g., Friedman, Hastie, and Tibshirani, 2007; Yuan and Lin, 2007), various regularization methods (e.g., Schafer and Strimmer, 2005; Li and Gui, 2006), and Bayesian methods (e.g., Madigan and York, 1995).”
MethodSparse and Locally Constant Gaussian Graphical Models
Honorio, J., Samaras, D., Nikos Paragios, A., Goldstein, R. Z., & Ortiz, L. E. (2009) · neural information processing systems · Cited by 21
Joint Estimation of Multiple Conditional Gaussian Graphical Models
Huang, F., Chen, S., & Huang, S.-J. (2017) · IEEE Transactions on Neural Networks and Learning Systems · doi:10.1109/tnnls.2017.2710090 · Cited by 21
“Yuan and Lin [7] and Banerjee et al.”
Variable selection in multivariate linear models with high-dimensional covariance matrix estimation
Perrot-Dockès, M., Lévy‐Leduc, C., Sansonnet, L., & Chiquet, J. (2018) · Journal of Multivariate Analysis · doi:10.1016/j.jmva.2018.02.006 · Cited by 21
“This technique has also been considered by [9], [10] and [11].”
MethodGraphical Lasso and Thresholding: Equivalence and Closed-form SolutionsPreprint
Fattahi, S., & Sojoudi, S. (2017) · arXiv (Cornell University) · doi:10.48550/arxiv.1708.09479 · Cited by 21
“There is a large body of literature suggesting that the solution of GL is a good estimate for the unknown graphical model, under a suitable choice of the regularization parameter [16, 17, 18, 19, 20, 21].”
BackgroundMethodJoint Learning of Multiple Differential Networks With Latent Variables
Ou-Yang, L., Zhang, X.-F., Zhao, X., Wang, D. D., Wang, F. L., Lei, B., & Yan, H. (2018) · IEEE Transactions on Cybernetics · doi:10.1109/tcyb.2018.2845838 · Cited by 23
“To learn the common supports of differential edges among K data sets, we introduce the following group lasso penalty [34]:”
MethodKronecker-structured covariance models for multiway data
Wang, Y., Sun, Z., Song, D., & Hero, A. O. (2022) · Statistics Surveys · doi:10.1214/22-ss139 · Cited by 20
“Besides being of interest in their own right, covariance estimates are often a prerequisite in other applications, including: dimensionality reduction; multivariate clustering; supervised classification; linear prediction; and model selection [42, 63, 101, 5, 24, 79, 80, 81].”
BackgroundMethodMulti-Scale Factor Analysis of High-Dimensional Functional Connectivity in Brain Networks
Ting, C.-M., Ombao, H., Salleh, S.-H., & Latif, A. Z. A. (2018) · IEEE Transactions on Network Science and Engineering · doi:10.1109/tnse.2018.2869862 · Cited by 24
“The third imposes sparsity on the precision matrix in graphical models with ‘1 penalization [37], [38], extended to a lowrank plus sparse estimation by combining latent variable”
MethodLinear operator‐based statistical analysis: A useful paradigm for big data
Li, B. (2017) · Canadian Journal of Statistics · doi:10.1002/cjs.11329 · Cited by 28
“Li, Chun, & Zhao (2014) used linear operators of the type described in Section 2 to extend the Gaussian graphical model (Meinshausen & Bühlmann, 2006; Yuan & Lin, 2007) to nonGaussian and nonlinear linear data.”
Feature selection for data integration with mixed multiview data
Baker, Y., Tang, T. M., & Allen, G. I. (2020) · The Annals of Applied Statistics · doi:10.1214/20-aoas1389 · Cited by 20
“A plethora of works have studied feature selection in the high-dimensional setting via the Lasso or GLM Lasso (Tibshirani, 1996; Yuan and Lin, 2007; Tibshirani et al., 2013; Zhao and Yu, 2006), and others have studied various data integration problems (Hall and Llinas, 1997; Shen, Olshen and…”
BackgroundUltrahigh dimensional precision matrix estimation via refitted cross validation
Wang, L., Chen, Z., Wang, C. D., & Li, R. (2019) · Journal of Econometrics · doi:10.1016/j.jeconom.2019.08.004 · Cited by 22
“Various penalty functions can be applied, such as ℓ1-penalty (Tibshirani, 1996), SCAD penalty (Fan and Li, 2001) and MCP penalty (Zhang, 2010).”
MethodNetwork Topology Inference From Heterogeneous Incomplete Graph Signals
Yang, X., Sheng, M., Yuan, Y., & Quek, T. Q. S. (2020) · IEEE Transactions on Signal Processing · doi:10.1109/tsp.2020.3039880 · Cited by 20
“Note that the precision matrix estimation methods such as the Graphical Lasso in [5] and [6] search for solutions that allow both negative and positive edge weights in the graph, while the work in [7] focuses on learning graph with non-negative edge weights, which restricts the precision matrix to…”
MethodA Normality Test for High-dimensional Data Based on the Nearest Neighbor Approach
Chen, H., & Yin, X. (2021) · Journal of the American Statistical Association · doi:10.1080/01621459.2021.1953507 · Cited by 20
“…T2-test), as well as many recently developed methodologies, such as network inference through Gaussian graphical models (Ma, Gong, and Bohnert 2007; Yuan and Lin 2007; Friedman, Hastie, and Tibshirani 2008; Rothman et al. 2008; Fan, Feng, and Wu 2009; Yuan 2010; Liu 2013; Xia, Cai, and Cai 2015),…”
MethodDistributed Lag Approximation to Linear Time-Invariant Systems
Robinson, P. M. (1979) · The Annals of Statistics · doi:10.1214/aos/1176344672 · Cited by 21
Accelerating Bayesian Structural Inference for Non-Decomposable Gaussian Graphical Models
Moghaddam, B., Emtiyaz, K., Murphy, K. P., & Marlin, B. M. (2009) · neural information processing systems · Cited by 33
Bayesian elastic net single index quantile regression
Alshaybawee, T., Midi, H., & Alhamzawi, R. (2016) · Journal of Applied Statistics · doi:10.1080/02664763.2016.1189515 · Cited by 20
Post-Regularization Inference for Time-Varying Nonparanormal Graphical ModelsPreprint
Lu, J., Kolar, M., & Liu, H. (2015) · arXiv (Cornell University) · doi:10.48550/arxiv.1512.08298 · Cited by 20
“Zhou et al. (2010) proposed to estimate the inverse correlation matrix by plugging Σ̂P into the graphical lasso (Yuan and Lin, 2007) objective Ω̂(2)(z) = argmax Ω 0 { log det Ω− tr ( ΩΣ̂P (z) ) − λ‖Ω‖1 } , resulting in the kernel graphical Lasso estimator.”
BackgroundGraph quilting: graphical model selection from partially observed covariancesPreprint
Vinci, G., Dasarathy, G., & Allen, G. I. (2019) · arXiv (Cornell University) · doi:10.48550/arxiv.1912.05573 · Cited by 23
“There has been an abundance of work on this problem [5, 46, 47, 65, 66], but we study a new version of the graph selection problem that arises when many pairs of variables are never jointly observed.”
BackgroundMethodInfluentialGraphical model selection and estimation for high dimensional tensor data
He, S., Yin, J., Li, H., & Wang, X. (2014) · Journal of Multivariate Analysis · doi:10.1016/j.jmva.2014.03.007 · Cited by 19
Learning Sparse Gaussian Graphical Models with Overlapping Blocks
Hosseini, M. J., & Lee, S. (2016) · neural information processing systems · Cited by 19
Matrix Graph Hypothesis Testing and Application in Brain Connectivity Alternation Detection
Yin, X., & Li, L. (2017) · Statistica Sinica · doi:10.5705/ss.202017.0023 · Cited by 27
“First, the key of graph estimation is to seek a bias-variance tradeoff, and many sparse graph estimators such as Yuan and Lin (2007); Friedman et al. (2008) are biased.”
BackgroundSparse Covariance Matrix Estimation by DCA-Based Algorithms
Phan, D. N., Thi, H. A. L., & Dinh, T. P. (2017) · Neural Computation · doi:10.1162/neco_a_01012 · Cited by 20
“…penalty ( 1-norm regularization) on the entries of the inverse covariance matrix to the normal likelihood (see, e.g., Meinshausen & Buhlmann, 2006; Yuan & Lin, 2007; Banerjee, El Ghaoui, & d’Aspremont, 2008; Friedman, Hastie, & Tibshirani, 2008; Rothman, Levina, & Zhu, 2008; Danaher, Wang, &…”
Property testing in high-dimensional Ising models
Neykov, M., & Liu, H. (2019) · The Annals of Statistics · doi:10.1214/18-aos1754 · Cited by 27
Estimating Sparse Precision Matrices from Data with Missing Values
Kolar, M., & P Xing, E. (2018) · KiltHub Repository · doi:10.1184/r1/6475607 · Cited by 19
Covariance-enhanced discriminant analysis
Xu, P., Zhu, J., Zhu, L., & Li, Y. (2014) · Biometrika · doi:10.1093/biomet/asu049 · Cited by 19
Sparse Machine Learning Methods for Understanding Large Text Corpora.
Ghaoui, L. E., Li, G., Duong, V.-A., Pham, V., Srivastava, A. N., & Bhaduri, K. (2011) · Cited by 19
Honorio, J., & Jaakkola, T. (2013) · arXiv (Cornell University) · doi:10.48550/arxiv.1309.6838 · Cited by 19
“…sparse regression for each variable (Mein-shausen & B¨uhlmann, 2006; Zhou et al., 2011), determinant maximization with linear inequality constraints (Yuan & Lin, 2007), a block coordinate descent method with quadratic programming (Banerjee et al., 2006; Banerjee et al., 2008) or sparse regression…”
MethodActive Learning Algorithms for Graphical Model SelectionPreprint
Dasarathy, G., Singh, A., Balcan, M.-F., & Park, J. H. (2016) · arXiv (Cornell University) · doi:10.48550/arxiv.1602.00354 · Cited by 19
“, [3], [4], [5], [6], [11], [28], [29], [30], [31].”
Inference for Sparse Conditional Precision MatricesPreprint
Wang, J., & Kolar, M. (2014) · arXiv (Cornell University) · doi:10.48550/arxiv.1412.7638 · Cited by 19
“We build on the work of Orabona et al. (2012) who developed a fast algorithm for finding the minimizer of the regular graphical lasso objective (Yuan and Lin, 2007, Banerjee et al., 2008, Friedman et al., 2008).”
MethodInfluentialGraphical model structure learning using L₁-regularization
Schmidt, M. (2010) · cIRcle (University of British Columbia) · doi:10.14288/1.0051935 · Cited by 114
Learning Graphs from Signal Observations under Smoothness Prior.Preprint
Dong, X., Thanou, D., Frossard, P., & Vandergheynst, P. (2014) · arXiv (Cornell University) · Cited by 19
Does the $\ell_1$-norm Learn a Sparse Graph under Laplacian Constrained Graphical Models?Preprint
Ying, J., Cardoso, J. V. D. M., & Palomar, D. P. (2020) · arXiv (Cornell University) · doi:10.48550/arxiv.2006.14925 · Cited by 19
“ived broad interest in statistical machine learning, where the problem can be formulated as a sparse precision matrix estimation. The authors Banerjee et al. [3], d’Aspremont et al. [9], Yuan and Lin [70] proposed the ‘ 1-norm penalized maximum likelihood estimation, also known as graphical lasso, to encourage sparsit…”
BackgroundMethodInfluentialZhang, L., Ma, R., Cai, T., & Li, H. (2020) · arXiv (Cornell University) · doi:10.48550/arxiv.2011.03598 · Cited by 19
“It is well known that the conditional dependence network can be modeled by Gaussian graphical model, which can be obtained using node-wise regression (Meinshausen and B¨uhlmann 2006; Yuan and Lin 2007).”
MethodHigh dimensional change point inference: Recent developments and extensions
Liu, B., Zhang, X., & Liu, Y. (2021) · Journal of Multivariate Analysis · doi:10.1016/j.jmva.2021.104833 · Cited by 37
Gene communities in co-expression networks across different tissues
Russell, M., Aqil, A., Saitou, M., Gökçümen, Ö., & Masuda, N. (2023) · PLoS Computational Biology · doi:10.1371/journal.pcbi.1011616 · Cited by 19
Bayesian Scalable Precision Factor Analysis for Gaussian Graphical Models
Chandra, N. K., Müller, P., & Sarkar, A. (2024) · Bayesian Analysis · doi:10.1214/24-ba1461 · Cited by 18
Chiquet, J., Rigaill, G. J., & Sundqvist, M. (2018) · Methods in molecular biology · doi:10.1007/978-1-4939-8882-2_6 · Cited by 18
Peterson, C. B., Osborne, N., Stingo, F. C., Bourgeat, P., Doecke, J. D., & Vannucci, M. (2020) · Biometrics · doi:10.1111/biom.13235 · Cited by 19
“vity networks. When all samples are collected under similar conditions or re ect a single type of disease, methods such as the graphical lasso 2 Biometrics, 000 0000 (Meinshausen and Buhlmann, 2006; Yuan and Lin, 2007; Friedman et al., 2008) or Bayesian network inference approaches (Roverato, 2002; Wang, 2012; Wang an…”
MethodLearning Dual Low-Rank Representation for Multi-Label Micro-Video Classification
Lu, W. D., Li, D., Nie, L., Jing, P., & Su, Y. (2021) · IEEE Transactions on Multimedia · doi:10.1109/tmm.2021.3121567 · Cited by 21
“(25) are L 1 regularized co-variance selection problems, for which efficient optimization was intensively studied in [47][48].”
BackgroundLocal linear estimation of covariance matrices via Cholesky decomposition
Chen, Z., & Leng, C. (2014) · Statistica Sinica · doi:10.5705/ss.2013.129 · Cited by 18
“Most importantly, we investigate the global convergence of our proposed conditional covariance matrix estimator under the Kullback-Leibler loss (Yuan & Lin, 2007; Levina et al., 2008) and the Frobenius loss (Bickel & Levina, 2008) in Theorem 3.”
MethodLeast angle regression for model selection
Zhang, H., & Zamar, R. H. (2014) · Wiley Interdisciplinary Reviews Computational Statistics · doi:10.1002/wics.1288 · Cited by 18
Large Gaussian Covariance Matrix Estimation With Markov Structures
Deng, X., & Yuan, M. (2009) · Journal of Computational and Graphical Statistics · doi:10.1198/jcgs.2009.07170 · Cited by 18
“This is the so-called graphGarrote estimator proposed by Yuan and Lin (2007).”
BackgroundMethodInfluentialSelf-citationStatistical Approaches for the Study of Cognitive and Brain Aging
Chen, H., Zhao, B., Cao, G., Porges, E. C., O’Shea, A. M., Woods, A. J., & Cohen, R. A. (2016) · Frontiers in Aging Neuroscience · doi:10.3389/fnagi.2016.00176 · Cited by 19
“Alternatively, graphical models are attractive for inferring brain connectivity due to their advantages over conventional marginal correlation based analysis (Lauritzen, 1996; Yuan and Lin, 2007; Koller and Friedman, 2009).”
BackgroundMethodInfluentialEstimating large covariance matrix with network topology for high-dimensional biomedical data
Chen, S., Kang, J., Xing, Y., Zhao, Y., & Milton, D. K. (2018) · Computational Statistics & Data Analysis · doi:10.1016/j.csda.2018.05.008 · Cited by 24
A Joint Graphical Model for Inferring Gene Networks Across Multiple Subpopulations and Data Types
Zhang, X.-F., Ou-Yang, L., Yan, T., Hu, X. T., & Yan, H. (2019) · IEEE Transactions on Cybernetics · doi:10.1109/tcyb.2019.2952711 · Cited by 21
“To reconstruct gene networks in high-dimensional cases where the number of genes is larger than the number of observations, various sparse Gaussian graphical models have been proposed [15], [16], [20], [21].”
BackgroundMethodEstimation and inference for precision matrices of nonstationary time series
Ding, X., & Zhou, Z. (2020) · The Annals of Statistics · doi:10.1214/19-aos1894 · Cited by 18
The Nonparanormal: Semiparametric Estimation of High Dimensional Undirected GraphsPreprint
Liu, H., Lafferty, J., & Wasserman, L. (2009) · arXiv (Cornell University) · doi:10.48550/arxiv.0903.0649 · Cited by 18
“Yuan and Lin (2007) and Banerjee et al. (2008) propose an estimator based on regularized maximum likelihood using an ℓ 1 constraint on the entries of Ω, and Friedman et al. (2007) develop an efficient algorithm for computing the estimator using a graphical version of the lasso.”
MethodAn empirical $G$-Wishart prior for sparse high-dimensional Gaussian graphical modelsPreprint
Liu, C., & Martin, R. (2019) · arXiv (Cornell University) · doi:10.48550/arxiv.1912.03807 · Cited by 18
“Under the sparsity assumption, regularization methods, such as graphical lasso and CLIME, are widely used; see Meinshausen and Buhlmann (2006), Yuan and Lin (2007), Huang et al. (2006), Friedman et al. (2008), and Cai et al. (2011).”
MethodA comprehensive survey of recent research on profile data analysis
Liu, P., Xu, H. T., & Zhang, C. (2024) · Journal of Quality Technology · doi:10.1080/00224065.2024.2369933 · Cited by 17
“…11 , ::: , x i 1 q , ::: , x ip 1 , ::: , x ipq � 0 2 R qp � 1 be the concatenated FPC score vector that follows a multivariate Gaussian distribution with covariance matrix R ¼ H − 1 2 R qp � qp : H is the precision matrix that can be estimated by applying graphical group LASSO (Yuan and Lin 2007).”
Conditional Functional Graphical Models
Lee, K., Ji, D., Li, L., Constable, R. T., & Zhao, H. (2021) · Journal of the American Statistical Association · doi:10.1080/01621459.2021.1924178 · Cited by 17
“Within this class, there is a rich literature on unconditional graphical models (Yuan and Lin 2007; Friedman et al. 2008; Peng et al. 2009, among others).”
BackgroundZhang, C., Zheng, B., & Tsung, F. (2023) · Data Mining and Knowledge Discovery · doi:10.1007/s10618-023-00916-w · Cited by 17
“In particular, Qiao et al. (2015) extended the graphical lasso (Yuan and Lin 2007) to multivariate functional data by first using functional PCA (FPCA) for feature extraction and then applying FPCA scores to graphical LASSO.”
MethodOn some algorithms for estimation in Gaussian graphical models
Højsgaard, S., & Lauritzen, S. L. (2024) · Biometrika · doi:10.1093/biomet/asae028 · Cited by 17
Estimating time-varying networks for high-dimensional time series
Jia, C., Li, D., Yu-ning, L., & Linton, O. (2025) · Journal of Econometrics · doi:10.1016/j.jeconom.2024.105941 · Cited by 17
Indirect multivariate response linear regression
Molstad, A. J., & Rothman, A. J. (2016) · Biometrika · doi:10.1093/biomet/asw034 · Cited by 17
An efficient algorithm for sparse inverse covariance matrix estimation based on dual formulation
Li, P., & Xiao, Y. (2018) · Computational Statistics & Data Analysis · doi:10.1016/j.csda.2018.07.011 · Cited by 17
Weighted Fused Pathway Graphical Lasso for Joint Estimation of Multiple Gene Networks
Wu, N., Huang, J. T., Zhang, X.-F., Ou-Yang, L., He, S., Zhu, Z., & Xie, W. (2019) · Frontiers in Genetics · doi:10.3389/fgene.2019.00623 · Cited by 17
“Gaussian graphical model (GGM) is an attractive paradigm to depict the associations among biomolecules (Yuan and Lin, 2007).”
BackgroundMethodFast and Separable Estimation in High-Dimensional Tensor Gaussian Graphical Models
Min, K., Mai, Q., & Zhang, X. (2021) · Journal of Computational and Graphical Statistics · doi:10.1080/10618600.2021.1938086 · Cited by 17
Metadata
- Authors
- Yuan, M., & Lin, Y.
- Affiliations
- Georgia Institute of Technology (US)
- University of Wisconsin–Madison (US)
- Venue
- Biometrika
- Publisher
- Oxford University Press (OUP)
- Published
- 2007-02-07
- Volume
- 94
- Issue
- 1
- Pages
- 19-35
- ISSN
- 0006-3444 (ISSN-L)
- 1464-3510
- OpenAlex
- W2081746825
- Semantic Scholar
- 20ee888ca9c3bae784f266405b4d6395c6f32955
- Cited by
- 1,627
- References
- 16
- Citations by year
- 2026: 43 · 2025: 66 · 2024: 60 · 2023: 64 · 2022: 86 · 2021: 131 · 2020: 112 · 2019: 147 · 2018: 142 · 2017: 91 · 2016: 123 · 2015: 98 · 2014: 109 · 2013: 86 · 2012: 85
- FWCI
- 38.99
- Type
- journal article
- Language
- en
- Open access
- closed
- Locations
- Biometrika (publishedVersion)
- RePEc: Research Papers in Economics (submittedVersion)
- HighWire Press Open Archive (submittedVersion)
- Funders
- National Science Foundation
- Topics
- Physical Sciences › Mathematics › Statistics and Probability › Statistical Methods and Inference
- Physical Sciences › Mathematics › Statistics and Probability › Random Matrices and Applications
- Physical Sciences › Computer Science › Artificial Intelligence › Bayesian Modeling and Causal Inference
- Keywords
- Gaussian graphical models, model selection, penalized likelihood, sparse estimation, shrinkage estimation, hyperparameter optimization, covariance structure
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.